Regarding the Gen 3 Egg Legality Issue

[papajefe]

I'd like to open a discussion regarding your recent decision on the flagging of Gen 3 Eggs with Method 1/2/4 PID/IV combinations as "Invalid" as opposed to "Fishy." (#3894 & #3092)

Before I get into my reasoning, I'd like to preface by saying that I have the utmost respect for what the development team of PkHex has created. As someone deeply invested in acquiring "statistically improbable" Pokemon on my physical cartridges without external save editing, I look to the PkHex green checkmark as a gold standard for achieving parity with what's possible for the games to generate.

Whether I'm using RNG manipulation, or abusing Pomeg Glitch mechanics to alter Pokemon's data on real hardware, my bar is set at "could this Pokemon have been generated via normal game functions."

I was not aware a duplicate issue had been submitted, nor did I submit this issue with intent to "prove a point." I encountered what I believed to be unexpected behavior, while trying to acquire a competitive Pokemon for my own use.

I'm aware this is a busy time for PkHex development as Pokemon HOME just released for Scarlet/Violet, so I'm conscious that this is not an urgent matter. Additionally, I know the relationship between the RNG community and the PkHex team isn't perfect.

However, my inquiry is coming from a place of seeking genuine understanding and agreement. I don't wish to stir up any animosity or make any enemies.

I only ask that you hear me out, and consider my thoughts on the matter before committing to your decision.

I realize I'm going to extreme lengths here to justify something that, as an RNG manipulator, I could simply ignore... but it's an issue that I think perfectly encapsulates a distinction in the true purpose of PkHex's legality checker.

It's not labeled as a "Hack Detection" system.
It's labeled as a "Legality Verification" system.

The difference between those two might be seemingly slight and semantic, but I feel it's significant enough to take a passionate stance on the issue.

First and foremost, the most important feature of "legality verification" is parity with in-game function. Presumably, this is the reason why PkHex has such a robust codebase with respect to the game's RNG systems.

To imply something is invalid based on statistical improbability opens the door to flagging an uncountable number of things as "Invalid" as opposed to what they are: "Fishy."

You stated that this flagging was preferred behavior.

"Preferred behavior" implies that this scenario is representative of a general philosophy with regards to legality verification and I'm curious about what exactly that philosophy might be.

From my perspective, PkHex is first and foremost a save editor. The primary use case is just that: editing Pokemon data.

I'm sure most of the greater community would agree that this software gives people the tools to "hack" Pokemon with their desired stats/natures/hidden power types etc...

However, obviously it's evolved into much more than that. Among other things, it's also a backup management system, a data preservation system, and a system for verifying if Pokemon received via trades/giveaways are legitimate or hacked.

This issue falls under that third umbrella, and from the reasoning given in the original issue, the general stance would seem to be that "hack detection" takes precedence over "in-game legality." This is where I believe our philosophy on the matter differs.

However statistically improbable this one specific edge case may be, it is not drastically different from many other statistically improbable Pokemon.

There is only one RNG state out of all ~4.2 billion Method 1 spreads that produces a Pokemon with a Timid Nature and 6 perfect IVs.

And yet any static/stationary gift with this PID/IV combination would be considered valid under the current system. This isn't even taking Emerald fixed seeds, Ruby/Sapphire dead batteries, or Shiny into consideration, which multiplies the magnitude of statistical improbability.

Nevertheless a Shiny Timid Method 1 Rayquaza with 31 in every stat isn't even flagged as "Fishy."

Even taken one step further, what if we instead used a Japanese copy of Emerald and put this combination on to a Shiny Mew event? From PkHex's current perspective this is seen as 100% valid, regardless of improbability.

It's entirely reasonable that given 20 years of gameplay and millions of players, someone somewhere has hatched an Egg with a valid Method 1/2/4 PID/IV combination.

In fact, at bare minimum there are 5 sets of IVs that could meet this criteria for any given PID, if you include Colosseum\XD and BACD_U (which I tested and are also flagged under the current implementation). That increases the odds of occurrence to 5-in-2^32. Five times more likely than that of my previous Timid 6IV Method 1 example.

The point being, taking a stance of simply: "it's statistically improbable to be legitimate, and thus this is the preferred behavior," doesn't seem to be applied consistently.

Even looking at it from the inverse perspective, let's say someone did try to hack a Pokemon in this way.

This would require they have enough knowledge of the game's RNG functions to grab a valid PID/IV combination, and then somehow lack to foresight to realize they could either:

A. Change any IV by 1 to achieve legality.
B. Change to any arbitrary PID to achieve legality.
C. Change Met Data to match a valid encounter.

Presumably they also would be using PkHex to do so, and given the way it currently functions, would be notified it's invalid and have reason to try any of the above options.

And so, marking these Eggs "Invalid" as opposed to "Fishy" under the pretense of "it's more likely to just be a bad hack" also doesn't seem consistent.

I truly adore this piece of software, and use it almost daily. I wouldn't be able to manage a collection of saves and Pokemon that date back nearly 20 years without it.

It might seem like a trivial footnote at the end of the day. Yes, I could just ignore this one tiny detail-- with full knowledge that I know the Pokemon in question are "legal" with respect to how the game functions.

But as I said before, I look to PkHex's legality as the "gold standard." I appreciate and respect the work Kurt puts into analyzing the teams of official competitions, and take pride in only using Pokemon that meet this standard.

In the interest of understanding the development team's philosophy on this matter, I hope you'll take the time to respond.

Thank You,

Papa Jefe

[Lusamine]

This is still not going to be changed, regardless of how many people keep making issues in different places, or how many different opinions you have. Allowing extremely unlikely edge cases where every single person has to RNG manipulate it to make the issue in the first place is more likely to allow bad hacks such as someone simply taking a wild Pokemon and turning it into an egg.

Every person who RNG manipulated it knows what they did and can ignore it. Flagging the much more likely bad hacks is a better service to the general user who has no idea what they're doing. In the same way, PKHeX does not honor the errors from RNG manipulating a past gen game to match a future game, and sending those Pokemon into the future game where they become "past Gen OT cannot be current handler" and where the past gen OT can gain friendship. That error occurs very commonly from people sharing data between games, or turning a Pokemon from one game into a Pokemon from another game, and is significantly more likely than accidentally rolling up with the same trainer info on 2 games.

If it is critically important to you that your personal Pokemon are not flagged, the program is open source and you can modify the legality checker however you want for personal use.

Please drop this discussion.

[papajefe]

bad hacks such as someone simply taking a wild Pokemon and turning it into an egg.

Is this something that is commonly done? I guess someone could try to hack Egg moves onto a non-bred Pokemon, then realize that makes it illegal, then check the "Is Egg" flag and try to hatch it, thinking it would solve the issue?

At the end of the day, I must respect your decision on the matter. I just wanted to understand why, and give my perspective in hopes of maybe reaching a compromise. If your priority is to flag potential hacks, more so than mirroring what's possible in the game, then that's your priority. I don't agree in this specific instance that its more unlikely than other potential hacks that are currently "valid" but as you said, if I really want to see the green checkmark I can just code it myself.

Thank you for reading, sorry for wasting your time.

[Lusamine]

Yes, it's extremely common for people to take an existing Pokemon and try to change it to something they want. The same error occurs if you take a wild Pokemon from SWSH and convert it to an egg. The check is better if it flags more cases of people trying to force one Pokemon into another.

Keep in mind that not a single person who played normally and encountered this problem has reported it, but multiple people have sat around thinking of how to do it through RNG manipulation and then argue passionately about it every time, so it gets dull.

Yes, many of my personal Pokemon also flag, but I can ignore it or modify it on my personal version because I know why. The target audience does not know why and it's better for them to get that guidance not to use the program this way.

[papajefe]

Understood, I wasn't aware this was a common issue. So it makes more sense to keep the flag in place. I hadn't considered the error is actually serving more as a "try to hack better" notification, rather than a "you've received a poorly hacked pokemon" warning.

I just stumbled upon this after going about my own business, unaware that this had been covered here many times over.

Maybe someday, someone actually will hatch one of these Eggs normally and make an issue. Its only a matter of time! 🤣